Research On Spatio-temporal Correlations Based WSN Data Compression

Currently,Wireless Sensor Network(WSN)is widely used in environmental monitoring,transportation,industrial production,agricultural industry,military and other areas.For example,the basic principle of environmental monitoring is to obtain information about the monitoring environment by deploying the sensor nodes in the monitoring area and making them communicate with each other frequently.However,dense nodes and frequent sampling result in a large amount of redundant data between the original data.Redundant data cannot improve the performance of the entire network,and also can induce lots of consumption on limited network energy,storage space and network bandwidth.Based on the above factors,how to efficiently explore the correlation between sensor data and effectively extend the lifetime of wireless sensor networks by compressing sensing data has become the focus of current research.Aiming at the existing problems of the current data compression scheme in network architecture,space-time correlation exploration,data reconstruction precision and so on,this dissertation based on the data exploration,combines with hierarchical network structure,zigzag scan and data sorting to solve the problem of low reconstruction accuracy.The main contributions of this dissertation include the following three aspects:1)We propose an adaptive spatial compression scheme based on zigzag scan.Aiming at the problem that the data reconstruction accuracy of current compression schemes is low,the adaptive spatial compression scheme(ASCS)avoids the faults of the mistakenly deleted important elements in the existing research schemes by deleting the smaller coefficients of transformation matrix in a targeted way,and in this way,ASCS obtains the higher reconstruction precision.Meanwhile,the compression rate varies with the sensor data correlation and the adaptive threshold.This feature makes the scheme has a strong adaptive characteristics.The final simulation results show that this compression scheme can achieve higher reconstruction precision under the same compression rate as compared with other similar spatial compression schemes.2)We propose an adaptive compression scheme based on data sorting.Aiming at the performance of ASCS can be improved further,an improved scheme is developed by introducing the data sorting operation.Based on the result of practical application,the spatial correlation among sensor data can be further explored by employing data sorting.This compression scheme sorts the original sensing data,and then the smaller coefficients in the transformation matrix are targeted for deletion.The simulation results show that the sorted data has a stronger correlation as compared with the unordered data,and the scheme has better reconstruction effect than the previous schemes.3)We propose a hierarchical adaptive space-time compression scheme.The previous two schemes only explore spatial correlation,but they don't explore temporal correlation.This part realizes the deep exploration of spatio-temporal correlations by combining discrete cosine transform and wavelet transform.First,this scheme combines the discrete cosine transform(DCT)with the adaptive threshold compression algorithm(ATCA)to expolre the temporal correlation between the sensory data.And then,the cluster head node adapts the discrete wavelet transform and ATCA to explore the spatial correlation between the temporal-compressed data.The simulation results show that after wavelet transform,the data has better agglomeration effect than the data don't,and the reconstruction error of the scheme is smaller than the other similar schemes.